Transcription
在实验室的设置中，我们展示了一些相对论行为。相对论动力学发生在光速。但我们可以使用量子“嗯”[音乐]量子比特和量子算法来模拟它们。现在，如果我大规模地进行，那么我就应该能够模拟一些[音乐]最奇怪、最复杂或最吸引人的事物，比如黑洞，或者我们称之为“你知道”[音乐]我的意思是，我不想使用“时间旅行”这样的词，以及其他一些事情，因为有，但当然，你知道“嗯”一切量子物理学定义的最[音乐]复杂的东西，我们对宇宙的理解的边缘，应该能够在某个时候模拟它们，而这正是我所看到的，可能在十年后，或者四五十年后，我的我们未来的几代人将用最复杂的[音乐]量子机器来完成。感谢您加入改变世界的行列。我是你们的主持人埃迪·伊维尔，今天我们请来了印度科学学院班加罗尔分校的 CM Chandhaker 教授。他的团队最近通过构建印度首个 6 量子比特光量子系统，取得了一项开创性的[音乐]里程碑。这一突破使印度跻身于在光量子计算领域领先的国家之列。这是一个利用光子的量子特性为下一代计算铺平道路的领域，在医学、人工智能、材料科学[音乐]及其他领域具有巨大的潜力。今天，我们将深入探讨光计算的科学、挑战和未来，以及印度的作用。所以教授，我真的很感激您抽出时间参加播客。现在，我对量子物理学一直着迷，我相信整个世界都是量子力学的，你知道我的意思是“嗯”，你，我，树木，空气，任何东西。所以它非常美丽，而且你越深入研究，它就越像一个兔子洞，你知道。我想从你开始，我是说，你知道是什么首先吸引你走向量子物理学和量子力学。如果你们能向我的听众解释一下什么是量子物理学和量子力学，以及你们多年来对量子物理学和量子力学的理解是如何演变的，那就太好了。>> 是的。嗯，非常感谢你“嗯”，你知道，为这次你正在主持的会议或采访。我的意思是，这是一个很好的机会，可以尝试谈论量子物理学，你知道，并对我们正在走向的技术有一个想法。所以，我想说，这是一段从高中物理学开始的旅程，物理学本身就让我着迷。试图理解我们周围的世界是如何运作的，就像你今天正确地开始说的那样，我们知道你周围的一切都是量子力学的，但这不是“你知道”我们作为小学生开始的时候，它就是你周围的东西，你只是想越来越多地了解它们，它是经典物理学，它是电磁理论，它是原子物理学，然后你就碰到了一个真正的底部，那里有不确定性和量子力学，它们都逻辑上与我们生活的世界联系在一起，这是一个非常简短的旅程。这简直就是大自然在你周围运作的方式。这让我着迷，对吧？深入了解事物如何在你周围运作。它可以是一支小小的“嗯”钢笔，流动的墨水，以及来自这个宇宙到达我们的光，并理解它们周围的动力学，当你开始观察它们时，它会打开很多挑战，然后你开始选择你喜欢的东西。你知道，就像你周围有很多东西一样。你开始喜欢某种颜色，然后你就会挑选它们。你开始收集它们，然后它们就成了你的游乐场。对。所以，对我来说，这是一段这样的旅程。它是物理学，从物理学开始学习“嗯”统计力学、电磁理论和天文学。你知道，我的意思是，我认为作为一个孩子，天文学也是每个人都会着迷的事情之一。你看到光。所以光是我一直以来都着迷的东西，然后你开始了解它们，学习一些数学，这就是量子力学的旅程。所以教授，我认为是 1801 年，当托马斯·杨“嗯”进行了那个实验，双缝实验，我认为这是最迷人的实验之一，但你仍然无法理解它，你知道我的意思是，发生了什么？所以，如果你能解释双缝实验以及这个实验对我们的世界意味着什么，那就太好了。>> 是的，我们可以回到铜的形式，对吧？所以，所以，我知道一切都是颗粒状的或粒子性质的，光本身最初被视为粒子，对吧？所以，首先，人们认为它是粒子，然后是波动理论，它出现了，因为人们可以看到一些干涉。所以，现在，这解释了很多经典光学，人们试图研究。所以，恩克实验的意思是，或者“你知道”量子图景中的双缝实验的意思是，你可以有两个缝隙，然后发送光。所以，当然，有很多内部细节，缝隙应该有特定的宽度。所以，现在，如果我发送光，那么它必须同时通过两条路径。如果它同时通过两条路径，那么一旦它出来，就应该有一些干涉，就像水通过一个狭缝然后发生干涉一样。但如果它是一个粒子，那么它必须通过其中一个缝隙，对吧？所以它不能同时通过它们。所以，如果它通过其中一个缝隙，那么你不应该看到干涉。所以，现在，如果我“嗯”做一个双缝实验，那么它应该清楚地告诉我光是粒子还是波，对吧？所以，这是一个明确的，你知道，我说哪种理论是正确的。这是一个思想实验，但然后，根据条件，进行了实验，或者“你知道”我们考虑了它，两者都会通过。不，当你试图看到，当你试图看到光子通过哪里，或者粒子通过哪里时，它就像波一样。但如果你试图看到它的轨迹，它通过的方式，那么它就像一个粒子。而且，根据观察者的“你知道”看待它的方式，你可以同时观察到波动行为或粒子行为。对。所以，这在很大程度上“你知道”被带到了我们生活的现实世界，因为那就是“你知道”我们理解的，那就是我们观察到的，我们必须接受它，或者我们必须提出一个替代理论。所以，这就是它的立场，对吧？如果你不接受你所观察到的或你试图看到的，那么我需要建立一个新的理论来解释它。所以，有没有什么你可以同时做的事情？这让很多人困惑了大约两到三“嗯”十年，在量子理论形成期间，对吧？所以，“嗯”，但是数学作为一种工具，试图提供帮助，但你假设每个粒子都伴随着某种“嗯”行为。现在，说它是一个振荡粒子或者别的什么，有时对我来说也是错的，因为我又为它们定义了一些边界条件。但我们所知道的是，我们知道我们开始的时候是什么，我们也知道我们观察到的结果。所以，现在我们需要建立可能发生的事情，一个逻辑结构。所以，也许中间时间我们测量某些东西，我们观察它们，我们的理论模型必须解释整个过程。所以，现在，已经进行了很多实验。现在，今天我们正在基于它构建技术。这意味着“嗯”，它是波粒二象性。我可以利用波动行为或粒子行为。如果我只取任何经典的东西，比如一个板球，我只能把它当作一个粒子来对待，我只能利用它作为粒子能给我的东西。如果我取水或者有波动行为的东西，我只能提取积极的东西，或者可以来自波动性质的资源。但如果我取一个量子力学行为的粒子，那么我可以同时选择两者。当我想要粒子行为时，我可以选择。当我想要波动行为时，我可以选择。所以，它作为一种双重资源来到我这里，对吧？这就是我们今天利用量子技术所做的。我能同时利用波动和粒子行为吗？当然，这 comes in a more intricate level. 但从更大的图景来看，这就是我们正在做的事情。而这正是将这两者结合在一起，这就是我们所说的量子资源，其中叠加、干涉、纠缠“嗯”是用于技术 Thus the origin for all those big resources which we talk about, right? So, so very interesting you know, because this experiment proved the wave particle duality and it's so very confusing because in the macro world things are fine but as soon as you go into the micro world things starts getting like slightly >> No, yeah, yeah. I try to put it in a slightly more different way. It's not about things are dramatically changing. It's just that there is a transition which we don't, we fail to understand when I know up to because at a microscopic scale wave behavior is not prominent, right? That's what I should, I would say, right? So, for example, uh, I have a, a simple high school physics where we study the velocity is proportional to, uh, the temperature, right? And there is also deep broadly wavelength where we try to see an atom or any particle. So, there is an wavelength associated with, uh, certain temperature or something. So, behavior associated with it. So, when you're trying to think of a bigger particle or like, you know, when the temperature is also very, very large, we, our wave nature becomes, you know, very insignificant. That's how we are seeing a particle behavior here. Now, the, the, the subject of when the, the quantum particle transfers to a classical behavior, there's a very, you know, there's no clear boundary line. That itself is a topic of interest. But of course, we do see both the ends very clearly, and there is a transition which is happening. So, when, when things are at the atomic scale, when I'm talking about electron, photon, or atoms, both wave and particle nature predominantly are, are there. But the moment the sizes, many, many atoms are put together. So, what happens is that there is the interactions between them. And there, so, there's a constructive, destructive, you know, behavior which happens, you know, between them. And then the wave nature diminishes. We don't see the behavior. It's almost like, you know, when I talk about, uh, when I'm seeing sea waves, there is a clear, if there's a pattern associated with it. I get a very prominent constructive and destructive patterns. Right? So, there's a waves which flows down. But the moment you have a lot of disordered, you know, interactions which is happening, then everything cancels up and you see like a smooth surface or something else, right? So, that's also part, that's how we, we start, we start, you know, seeing the particle behavior very prominently, and you go to the classical, you know, bigger scale. And wave behavior is very insignificant. We can't predict it. >> So, what, what you're trying to say is there somewhere in the middle there's some kind of a transition which is happening which >> Yeah. So, what we call as classical to quantum transition is a, a topic of its own interest where we see, uh, there is a noise associated with it. And we make the, the world interact. And whether that is what is destroying the quantum behavior. When I say destruction of quantum behavior, which means that we are trying to, uh, diminish those wave behavior and trying to get to, uh, particle nature which you see in the classical world. Right? Right. No. Now professor, in, in the quantum, quantum world, there are a lot of things which is happening which one might call a term weird. Like, for example, the quantum superposition and the entanglement. For someone who's new in the field, will I mean, who doesn't understand this, I mean, can you kind of explain it and break it down like what quantum superposition and what quantum entanglement is and how does this play a role in the quantum >> Yeah. Uh, is a word which probably I commonly use, but for us, it's a very, you know, fascinated picture of the world, you know, the amazing feature that you see in this whole universe, right? So, okay, superposition in a text. Okay, let me first put it as a textbook definition point and then let's try to understand, right? You can have a single particle, either it's a photon or an electron or anything which we call a subatomic, exist in two places at the same time, right? Now, that is what we call as a superposition. Simultaneously, it can be in two places. But two is the most simplest way to put it. When we are trying to exploit it at the larger scale today as a technology, it can exist in superposition at multiple places. Right? So, that's what we call it a superposition. Now, how we can understand superposition at a, as a simple level is that now we have to bring in a wave behavior. Right? When we talk about the, the quantum at a, at a smaller level, they have a wave behavior, which means that now wave is spread around the given position space. Now, how much it spreads depends on the environment around it. Right? So, it can be like something can be destructive. You may not be able to spread it for a longer distance. It's almost like, you know, expanding your bubble gum, right? Or like, you know, you have some elastic. You need to expand. Just that, you know, the particle can be spread in superposition. Now, the moment I look at it, then it collapses. It's almost like you're trying to force something, you're trying to intervene to see where it is, then it's like, you know, when you say mercury was like a ball, and it's just at such one particular place. So, it ends up being at that place. So, you can make it spread. The moment you're trying to observe, it's gets identified. So, this is what we call as a superposition. Now, you can control how you can, you know, you can bring them in superposition state. This is to do with a single particle. So, can one particle be at two positions simultaneously? It's a very big puzzle which people are trying to look at. Even today, there are certain questions which, you know, at the foundational level of quantum mechanics, to try to address how much non-local it is. Like, you know, can a single particle be in two different places? And how do we clearly establish that they are? And what are the, the full, what you call proof measurement or methods can be used for it? So, uh, now that's, uh, still a topic of interest, but as a, you know, as an established one, they're very clear that you can have them in superposition in multiple places at the subatomic level as single particles. Now, entanglement is a, a next level of it. So, which would mean that can two particles be in superposition state into? And that's what it would mean like, you know, now, which means that I can have two particles and I create some kind of superposition of those two particles and then send them into two different places. Now, uh, you know, it can be some properties. It's not like, you know, I, I take a property like polarization or something like that. If it's a photon, or if it's electron spin. So, then I can take them apart. Because when I'm trying to create them in superposition, they would have some interactions. When I move them apart, the state of one of them would end up telling what is the state of another one. Like, you, I don't have to even make any, I don't even have to ask what it was. Now, this is like violation of Einstein's theory of relativity, speed of light transport, right? That's how Einstein did not want to believe something like this really existed. He called there's some spooky action happening there. There's something hidden, uh, variable which is really doing it. But it took a lot of time for people to, you know, come up with an understanding of it and, uh, do experiments and verify. But, uh, but that, that's, that's the reality. So, you can actually, but, but is it easily creatable? No, it's not. You can't create entanglement very easily. There are a lot of, it's very hard. But you can, you know, bring two photons together or, you know, or like you can actually generate them in some specific manner. Such that they are entangled. Then you can take them away for thousands of people actually dying for thousands of kilometers and then measure polarization in one of them and instantaneously say the other one should have been a different polarization. So, it's like instantaneous information exchange. And that's what entanglement is, is about. And, uh, every lab in our lab, we keep generating entangled photon pairs on a regular basis. And every time you do experiments, you know that you're seeing it happening in front of your eyes, you're measuring, and then you have to believe that's how, you know, uh, nature can, what that's what nature has to offer to us. And that's what is helping us to build resources. I think in between these two, a superposition and entanglement, which is there, two particles, there is an interference which comes into an important role, which is trying to help us to create a transition of, you know, the single particle superposition and many particles interference, or even single particle can interfere. That's there. Because it's what we see, what we explained about double slit experiment. It's like a single particle which goes, because it's a wave behavior, it's actually takes both the paths and then when they come out, there's an interference there of the single particle. Right? So, now I can, now I can even think of many slits. Right? First is two slits. The next one can be a four slit. At the next one can be a six slit. Now, the, the one particle which is actually taken two parts, the next time it ends up taking all those four parts in superposition. It can end up giving a, a completely, you know, a different form of interference on D. Now, that is what we use as a reference points to build our algorithms today. Quantum algorithms are a kind of, you know, such kind of interference behavior. Now, there can be constructive if you can engineer how the slits are made, right? I can make some slits missing, and I can have some slits slightly wider, some slits. I can engineer my slits. If I engineer my slits, then I can define where to have a maximum interference, constructive interference, where to have a, a destructive, okay, no interference at all, right? So, I can engineer them. Now, that is actually my algorithms. The where I get my constructive interference is my solution to the problem. That is how we do quantum computation itself. It's almost like engineering your paths for photons or electrons or qubits, whatever it is called. Superposition is what is defining our qubits. So, when I talk about, when we talked about superposition being in two, uh, slits together, is like having two spins up and down simultaneously together. And that is what is mathematically what we call as qubit. So, engineering superposition of, but as a single particle would be like controlling my qubits. Now, that is how we map the fundamental physics. A simple double slit experiment becoming multi-slits, controlling their dynamics is almost a map to the way quantum computers are engineered at a, at a very elementary level. Now, I can have many particles. I can have large number of slits. I can make them interfere and engineer them in different ways. And that leads into a very complicated problem. And that the solution what I measure is my solution to the problem. It can be a computational problem. >> I want to get into the quantum computing side, but before that, uh, the current computation paradigm, you know, I mean, that's been is the binaries, you know, which which uses zeros and ones. And for this, current computation paradigm, what we use are these silicon chips, you know, the, the silicon chips which are powered by electrons, you know, doing all the computation for passing through these wires. Now, you know, there's there's a saying that, you know, that Moore's law is reaching an end, you know, because now we kind of, you know, pushing around, uh, and and creating these chips which is, you know, beyond, uh, like the two nanometers, you know, I mean, there are these two nanometer chips that we have created. Eventually, I mean, are the quantum effects going to play, uh, play, and then maybe it's going to be the end of the Moore's law? What are your thoughts on that? I mean, is that current computation paradigm going to end because we won't have any space to kind of, you know, engineer more transistors into this chips? Uh, that, that, that's the first, first question. And then the second is, is photonics the future? Will photonics, uh, uh, play a role? And will also be great if you could maybe explain, you know, photonics and >> Yeah. I, I'll try to answer the, the first question. And I said, uh, so, was not an experimental, you know, um, obtained conclusions, but it was an observation over years how your transistor sizes are changing and how your computational powers or how the transistors we're putting. And, uh, that clearly said that if you're trying to double the transistors in your chips which you are building or in your silicon wafer, it reaches a stage where you are getting into a nanometer scale. Uh, it is, which when one, you know, in the beginning, what I tried to say was that if things are bigger, your particle behavior is very, very prominent. The moment when it's becoming smaller and smaller, wave nature also gets prominently visible. Now, that's what exactly is what Moore's law is also trying to, you know, take us to. Like, you reach a stage where you have to put your transistors or your operational, uh, classical bits devices at a size where wave behavior is getting prominent. The moment wave behavior is prominent, then first one to the next one, there is a, you know, spread of the waves. So, the moment you have spread of the waves, your classical bits or your information operations, everything gets mixed out. So, now you need to, you know, maintain a size where you are not getting to a limit where wave behavior starts entering into your picture. That means quantum mechanics starts coming to play a role. So, that's what it said as a limit. And then they said, if you want to go smaller than that, then you need to operate at a quantum mechanical level. Uh, or else you can't, you know, you should stop thinking going beyond that. That's what is most of. But, uh, I think, uh, these were all an initial ideas which were put out to say why one has to start thinking in terms of quantum dynamics. But I think much more fundamental here is that, you know, why we do computation or why we do anything at all, right? It's a binary computer was a language which was developed for us to make things easier, right? So, uh, might, or to make it automated, or like to get something 2x5, I know, I can just write it as 10. 2, sorry, 2 into 5 is 10 for me, straight away. But if you ask a computer to do it, it converts into binary terms and then it adds it and it gives me. But it's still a good language, uh, to begin with. And, uh, now this was a mathematical evolution which happened for us, and we got a tool. But now, if you see how nature is doing all these things for us, certain things are very naturally happening in nature, much more better than the way the computers do it. It's because nature is much more efficient. You know, what are laws of nature? It, if, if we all agree that we, none of us have come up with any law better than quantum mechanics today to explain what is happening around the world. And it's hundreds of years, 100 years now. Uh, every experiment and everything which has been done has only made sure that is correct. So, now you need, if you use laws of nature, and might be we'll be able to do far, far beyond than what we even think of today. Right? Now, today, we might be talking about few things what quantum computers can do. But, uh, if we are trying to understand nature and mimic nature to the best of it, it's possible, then, uh, what can open to us might be much more wonderful than what we think about, right? So, I think that's probably is much more, you know, important than Moore's law. Moore's law is a very engineering statement to say why, you know, we are to go for a quantum computer. But whether it's, it's not about economics for me. I mean, why, if you ask me why I'm interested in quantum computing, or why is quantum nature doing that? It's to understand nature. Let me try to do. And that itself is in fact a Feynman statement also. Inviting people to do quantum computing was that, see, if nature is doing, you know, or if I want to even mimic nature, I need to follow its rule. Only then I can mimic, or I will only be able to do some approximation of it. And that's what I see is a driving force for most of the physicists who are working in quantum computing or like, you know, quantum information as such. That's the, the first part of it. Now, I think the second one is a much more, uh, interesting question, which I may not give you a very, you know, an answer which would say as, yes, photonics is the future. Right? No, I'm not going to say, you know, because we don't know. Because there, there are two things. One, we definitely know, uh, quantum theory is what is defining everything. And if you look at the current scenario across the world, uh, any researchers, people are working on all possible platforms to build quantum computers. When I say platform, which means that when I mentioned about superposition or entanglement or interference, now, which are the natural or like, you know, nature's, uh, tool which is helping us to control these three features, right? Like, you know, it can now be a superconducting qubits, which is one of the, the promising, uh, candidate where you can actually create superposition. You can see some interference. You can also have entanglement. So, which means that you can actually do some computations with there. Photons are there, uh, which is, which is fantastic, uh, you know, uh, you know, gift from nature for us to play around with. And, uh, we have, uh, atoms. People have for decades, people have learned to cool atoms, trap atoms in one particular position, control their quantum behaviors. Ions are there. These are all, you know, clearly, uh, there about six of them which is now very, very, uh, uh, an important, uh, candidates for qubits. Now, every, all of them have their own advantages and a big disadvantages. There, there's a, there is a big hurdle which each and every hardware has to pass to have our dream machine built, right? So, we don't know. I mean, as a researchers, for us, we know what all are the positives, and we also know what is the, the hurdle which has to be crossed. And we believe that we, we probably can jump that hurdle and there will be some solution to overcome that hurdle. It's true with everything. I can just give, in the interest of time, I will just give an example for two of the, the, the prominent ones these days, like one, uh, you know, photonics. Uh, the positive thing is that it doesn't interact with the environment. So, it means that you can actually, you know, make them go, control how photons can move around. Wonderful tool. But because they don't interact, they don't interact with other photons also, you know, it does not help me to control, uh, the, the, the gate operation, what we call, you know, computation means you need to do some interaction operations. You can't make them interact the way you want. And that is one of the hurdles. Uh, or else everything is fantastic. I can generate photons, I can, you know, make them way, move around the way I want, I can measure them. It can be done in room temperature. But how to make, uh, interactions between them is, is one of the hard problems to be looked at. Now, if I look at superconducting qubits, they are, you know, not at the single atom level. It's a microscopic atom. You can create superposition, you can do some kind of impact. But now the problem with them is like, I can, I have to now keep all these qubits built. And then now all these qubits are a, a microscopic item. So, they generate their own noise. And those noise, the moment I'm having some few hundreds of qubits built, uh, is making the system lose the quantum behavior. It's decoing. That I can actually, I don't think, you know, it's, it's very hard task. You can have thousands of engineers and build few thousands of qubits of superconducting system, you can build. Now, can you make them work coherently without having a noise? Is a big challenge. It's a scientific challenge there. It's not just about a technological or a translational problem. We know, uh, we are aware of it. Researchers who work on it are aware of these problems. But now, only when we have problems, we can solve them, right? And there is no reason why one can't solve again. There is no any law or theorem which says there is no go beyond it. Right? So, it's, it's, it's a task on us to find out solutions for them. So, having said that, uh, I'm going to, uh, say that there is no clear, uh, winner right now. All of them are very good. And for next, I mean, I, I would, I strongly believe that for next one, two decades, all of them will be a very important players. I would say that we are creating a massive knowledge base right now. And probably that's going to be for next decade, uh, from an academic point of view. >> Right. Right. Professor, uh, thank you for explaining it so beautifully. I think we're living in such a fantastic point of time, you know, I think the, the current classical, uh, computing paradigm, it's completely upended our world, you know, and the kind of applications that it's giving, given us, it's like, it, it's super awesome. It's completely changed a world. And here we are at the cusp of maybe transitioning to a different paradigm. And, and you rightfully pointed out, you know, I think nature is quantum mechanical. And there we are going over there and possibly kind of, you know, trying to build a, a computer which plays with the source code of life, you know, atoms itself, you know. And when we are able to build, uh, machines, you know, quantum mechanicals, you know, whatever approaches you, you, you take, you know, whether superconducting qubits, ions, or photonics, I mean, you know, when we kind of, you know, do that and build something, I, I, I think, you know, it'll be really good. And, and you rightfully pointed out, I think each of these, uh, approaches has got its own advantages and, and its own disadvantages also. But I'm sure like, I think, you know, we are in this cusp of like a fantastic point of time. I think in the next few years, maybe this current paradigm, you know, will possibly help us maybe discover new materials. Maybe the current paradigm itself, you know, with artificial intelligence, maybe it'll help us maybe understand nature deeply. And maybe we'll be able to kind of simulate and build, build these machines, you know. So, fantastic. I, I think, uh, time that we are sitting in. And I think you're at the cusp of, you know, building something really, really cool, you know, because yes, I mean, you know, the six qubit quantum computing, I mean, the, the breakthrough. So, I would love to learn, understand more about when and how did you decide to build a photonic quantum computer in India? And what were the hardware and software challenges in building India's first 6Q, uh, photonic quantum, uh, computing system? Stick photonic quantum computing, there are many approaches. One of the approaches called continuous variable. The, the entanglement in, uh, in that approach was formulated in India, you know, there's geometric phases, which is also one of the important thing which evolved as quantum phases, and everything had some foundational work done in India. So, there has been a prominent contribution in theory. So, experiments, it happened at some point, uh, with some government meetings where they're talking about the possibility of quantum technology. What all had happened there? A lot of proposals. They tried to pick few people. There were very few people who, who had some background in the field. So, that's how I got involved at some point. And it was an, I always liked light and photons. And, uh, I said, okay, this is something which has not been touched. There is nobody who is trying to look at photonic quantum computing. And, um, that's how I said, okay, let's, uh, begin with. Not as a computing was not the first thing. It was almost like, um, generating entangled photons, understanding them, and eventually build it to quantum. But now, if you say the six qubit, which has been, uh, or like, it's not, I mean, it's okay. The use of word of first in India, first in all these things, it's probably good for a tagline. And for us, yeah, so, but as a researchers, we need to know what really it is, right? So, you know, there are certain global standards. But what we have also done is not nothing anything, nothing inferior, or it's not just that we are trying to catch up. But the theory for that, the model has been taken many years. It didn't happen at overnight. From 2009, there are some problems which we were trying to work out. How we can have universal quantum computation using, uh, a quantum walk model. And there's been, I think we wrote some papers in 2018, 16, 17, came up with mathematical model to have, uh, you know, because I said photons can't interact, uh, with each other easily. So, there, can I use different degrees of freedom of photons to make them interact and make gate operations, CCl? So, this, there was theoretical model which we had developed, mathematical framework for it, and say, mathematically, it is possible. Then, uh, those papers were, uh, uh, you know, published. Couple of papers, uh, were published to say, look at, uh, the physical, practical feasibility. But then, how one has to do experimentally, and how I can, you know, control photons' degrees of freedom, that we started working on after we got, uh, the funding to set up a lab at IA. And, uh, we step by step, we started, uh, converting the, the mathematical model which we had developed at Instead of Mathematical Sciences, uh, so with my students, I developed it. And then we converted them into, uh, an optical scheme in IA. And we executed them step by step. So, it's a long years. It's not just over like, you know, might be some of them say, oh, you, you set up a lab and very quickly, you got something. No, okay, because we had clarity from many years of our work, you know, my, there were two PhD students who had graduated building up a mathematical model. And there are two more PhD students who started entering into this. And there were many few postdocs who worked on optimizing the scheme. And then we, we implemented it over again. Is it when we talk towards six qubit? What it can do is again, a very important question, right? I know we have generated a six qubit entangled state. And we have implemented few gates. But that doesn't mean that we can do anything with with that. So, there is the number of gate operations limitation right now. And we are trying to improve on. So, the next challenges are for us to increase the number of qubits plus the number of gate operations we can do. So, to address some of the, the problems, u, this process >> Right. Right. Yeah. So, I, I think, I mean, the, the general narrative that goes out to the world is that quantum computer is going to completely change the world. It's, it's the fastest, uh, thing in the world. But, you know, when you understand deeply, then you realize that, okay, quantum computing has its own problems. And, and it, it's, it's got a narrow use case at this point in time. Eventually, if we are able to kind of create, uh, error, uh, error-free quantum computation, yes, I think then maybe we will start scaling. And then maybe there'll be like, you know, two applications that will be coming out, uh, uh, from there. So, so my first question is, professor, I mean, you know, where do you currently stand with this quantum photonic system? And is this architecture scalable, you know, right now, you mentioned that, you know, this is like a six qubit thing where you've encoded three qubits per photon, uh, using the Greenberger-Horne-Zeilinger state. Uh, where does this stand at at this point of time? Is this architecture scalable? Can we scale this? >> So, I, I think there's, there's two questions I would like to answer. Probably one was not asked as an explicit question, but that's a very important question which I want to answer is that, uh, you know, when we see what is the future, whether quantum computing is going to solve X, Y, Z problems, these are all like, very, you know, the questions may not even be relevant at this point. Because we all try to think of the possibilities depending on the problems which you see as problems. So, I just have a very historic, why sometimes it still helps us, like, you know, around the beginning of the previous century, not now, in the 1900s, people had put up a challenge, how, how the technology is going to change, you know, visualize what it's going to be next 20 or 30, 40 years. And then when people look at after 40, 50 years, everything was wrong. We were far beyond then what it was. A lot of things which were not even expected. People never thought about it was achieved. It's left to us, uh, a creative minds, which is there. There are plenty of them. You know, our generation are trying to like, our generation are probably who are exposing, are trying to see things in very pocketed way, right? The generation which are very, very young now, today, 15, 20 years, they have grown up, they've seen mobiles, they've seen technologies, is beginning with that. For them to even to explain how the world was when we are not able to reach, when we are not able, it's, it's very, very difficult. But their thinking and their imagination on what they want for their future is not easy for us to even imagine. Because my, my trained mindset narrows down my thought process. My trained mindset narrows down what I have to visualize. So, we, we might try to put six, eight big things, what computers can, quantum computers can do. But I would just leave that open. Uh, our job is to, as I said, create knowledge base, leave it for the next generation to take and let them imagine and build on it. As long as it is honest, correct setup which you have built, that's going to take over and look at it. So, that's why we should not stop just, you know, doing it in technology. The science and technology has to go in, go hand in hand, and create a knowledge base. That's first. Now, I think the most important question coming to our, uh, research or architecture, yes, when I said, when we were trying to build a mathematical model, the two important things, yes, we had to, uh, address the issue of interaction, because photonics gate operations were probabilistic, because with certain probability, they used to be successful. And we wanted to increase that to some extent that we could do it at a three and six qubit level. Now, if I want to increase the number of qubits, then I need to involve more and more photons. Then it does become a little bit, it would just become probabilistic, but it will be better than any existing schemes, probability behavior. So, which means that it would be less probable than any known schemes. So, in that way, I would say it is better, uh, than the all the known existing, uh, photonic discrete variable photonic computation model. So, scalability, yes, it's scalable, because that's what when we say, uh, if I, if it's not scalable, then there's no point for us to build a mathematical model. Then I know, I should say, we have built some, uh, four qubit or six qubit system, which is useful for some communication or cryptographic applications, because smaller number of qubits have some applications there, and they are still a useful tool, and I should stop there. But when we are using a word, uh, computing, then it means that we do have a scalable architecture, uh, up to, I, in fact, we have kind of worked out few up to 100 something, and then it means that we can scale it up, uh, beyond that. How do we scale? Uh, there are multiple ways. Right now, uh, in our, uh, purview, we are trying to use a combination of, uh, miniaturized bulk optics and try to scale up to 30, 40, to 50 qubit level and demonstrate few things. Because it's not just about the qubits, and I can actually have lot of photons, and I can say I have large number of qubits. I can also do that with photons. But that does not mean that, you know, I can do some operation. So, we took, now, what, what is our task for next year or two is that, uh, while scaling, demonstrate few, uh, algorithms implementation, few sample cases, that gives a clear architecture. Because, you know, when we mathematically build certain architecture, uh, then we say, okay, this is how it works. When we try to build an experimental setup, then there will be some corrections to the mathematical model. And that we have already done at a six qubit level. When I have to take it to 12 qubit, next target for us is to go to 12 qubit. When I'm going for a 12 qubit level, then I might have to make some modifications to the scale to make it easier. So, when I go for a next level, and these improvisations and modifications will help us to take it to integrated circuits later on, which becomes, then I don't, we don't have to manually build such kind of thing. There can be a mechanized process of having integrated circuits built in a fabrication facility. And then scalable, scalability won't become a very difficult task. So, again, as I said, we have to build now, a knowledge base, uh, to fix some minor issues, errors, or corrections to the existing model. It would more or less be improving and making things better. >> Right. Right. So, so professor, I mean, yes, I think you already mentioned that, I mean, you know, when we get like a working quantum computer, the applications might be something that we are unable to imagine at this point of time. But will be great if you could, you know, maybe, you know, give some kind of understanding, you know, when you say that, you know, maybe when you scale from 6 to 12, uh, uh, qubits, when you say that you want to demonstrate something, maybe it would be great if you could kind of give us a glimpse of like, what could be those demonstrations? What do you see the, the, the, the near field, maybe terms of applications? And what could be, will also be great if you could like, you know, push beyond and, uh, maybe imagine and give the audience like, what could be those, you know, far out, uh, uh, far-term applications of quantum computing that you're very excited about? >> What we have already shown with a smaller number of qubits that kind of helps you to say like, you know, about 12 qubits or like, you know, other when we scale up, what we, we imagine to do. So, uh, a simple two qubit system, what is it useful for? Right? We can say, and I have a two qubit system. Why is it
你知道，一个经典的，你知道，一个经典系统可以做很多事情。两个量子比特有什么用处呢？所以我们用两个量子比特系统做的是构建一个量子随机数生成器。我们实际上构建了一个设备，目前该设备已被一些政府组织用于密钥生成。那么，这个密钥生成是什么呢？为什么量子力学能帮助你呢？这是一个很大的话题，但我会把它讲得很短。今天我们获取我们的 OTPs，今天我们所有人都依赖于一些我们即将获得的 PIN 码。这些 PIN 码是如何生成的？现在有一个软件会创建 PIN 码，然后他们将其提供给银行。现在，这个软件是一个已经编写好的程序。现在，这个已经编写好的程序实际上是可以被操纵的。有人实际上可以扮演一个恶意的角色，给你你想要的 PIN 码。但是，现在如果我要求大自然给我数字，大自然给我 PIN 码，那么这是最不可控的方式。当我们谈论波粒二象性时，一开始就有波包的坍缩，它会扩散，但只有当我看着它时，它才落在这个点上，否则就不会。所以我甚至不能说，如果我构建这个设备，我不能说它会坍缩在哪里。它在一个特定的点上坍缩。即使我构建了一个设备并把它给你，它也不是在我编程的层面上坍缩的。是大自然在决定。现在我们所做的是，我们采用了一个两个量子比特的系统。现在两个量子比特的状态会是已知的，我现在可以将它们转换成简单的二进制语言。我可以得到 0 0 1 1 0 1 1 这些可能性。现在当我们说它坍缩到其中一个时，这意味着我瞬间得到了 0 1 或 0 1 0 1 1 中的任何一个，我无法控制，大自然在提供这些数字，然后这些数字被转换成 OTPs 或捕获码或 QR 码或一个捕获码或 QR 码，这就是我们用两个量子比特系统所做的，目前它已被政府机构用于他们的应用程序。所以这是第一个，你知道，这就是我们如何尝试测试可扩展性。现在当我们转向 4 个量子比特，6 个量子比特时，但去年我们有 4 个量子比特，并且我们用它展示了其他东西，我们谈论公钥和私钥，以及图像的加密和解密，所以我们因为你知道，当光子可以纠缠时，我可以取一个四量子比特系统，两个量子比特被发送到一个点，另外两个量子比特被发送到另一个地方，现在这实际上可以被视为公钥和私钥。现在大自然在提供这些，我不需要做任何数学运算。你知道的，当前密码学的自然方式，数学扮演着重要角色，这些密钥是由复杂的数学问题生成的，以及破解的难度，你知道的，分解一个大数，你有一个大数，你必须分解它，一个给一方，另一个只有当它们匹配时，你才在决定。现在这是一个数学上困难的问题，它由数学定义。所以现在人们已经尝试提出非常非常复杂的数学复杂问题，我们可以将它们用于密码学，但有了量子物理学，大自然只是简单地给了我一个四量子比特系统，它复制了这种行为，这就是我们构建的设备之一，比如我们目前正在原型化它，因为光子在室温下工作，我们实际上将两个量子比特设备运送到了不同的地方，它去了德里，它去了其他地方，它已经被使用了，现在我们正在制造一个完整的紧凑型原型设备，它可以为用户生成公钥和私钥。所以我只是想说，你知道，就像这样，有简单的四个、五个、六个量子比特的即时应用，因为最终，如果政府在研发上投资公共资金，我们也必须表明有一些即时用途，因为长期用途我们可能会想到很多事情，并且可能会出现其他事情，但即时用途非常非常重要，而这些就是其中一些要点，但现在，当我们转向 6 到 8 个量子比特时，有一些简单的自然动力学，可以是简单的化学复合物相互作用或我们想明确建模和模拟的东西。经典计算机可能会解决这个问题，但如果你表明我们实际上也可以定义一个量子算法，那么它就证明了这个算法是有效的，然后人们可以为更大更复杂的复合物构建算法，并继续观察它。所以，你知道，当我们从更少的量子比特扩展到更多的量子比特时，我们正在设想这样的问题。最清晰和已完成的是安全应用，在较少量子比特的数量上有一个明显的赢家，我认为这个和光子学是为此量身定制的，其他硬件，你知道，你无法携带低温超导量子比特系统来做到这一点，而且它也非常，但是这个光子系统和光子传播，你不需要像超导量子比特那样固定在物理位置上，它不会传播，所以光子量子比特实际上可以传播，所以我可以在那里创建量子比特，我可以将它们发送到不同的地方，然后我看着它。所以，这就是为什么我说，你知道，光子学，你无法将光从你的量子世界图景中排除。光以某种方式存在，并且它将发挥作用，因为它帮助我们在不同的地方移动。这些是一些小规模的应用，但当涉及到大的想象力问题时，我不会谈论行业对此的看法，经济世界对此的看法。对我来说，我个人的热情一直是模拟大自然，大自然中复杂的行为。当你问我是否可以用你的想象力时，对我来说，即使是黑洞也是一种想象，对吧？如果我的物理学是正确的，如果量子力学帮助我们理解黑洞领域发生了什么，那么我应该能够在实验室环境中模仿它们，对吧？所以有一些事情，事实上，我与马里兰大学合作的一个有趣的问题，我们称之为德拉方程的较小的离子阱系统，它是一种相对动力学，我开发了一个算法，而不是数学科学。我们与马里兰大学合作。今天，它已经发展成为 Ion Q 量子计算公司，一家大公司，但通过它，我们实现了这个算法，在实验室设置中，我们展示了一些相对行为，相对动力学和光速，但我们可以用量子比特和量子算法来模仿它们。现在，如果我大规模地做到这一点，那么我就应该能够模仿一些最奇怪、最复杂或最吸引人的事物，比如黑洞，或者我们称之为，我的意思是，我不想使用像时间旅行这样的词，以及其他一些事情，因为有，但当然，你知道，量子物理学定义的一切，我们对宇宙理解的最复杂的前沿，你应该能够在某个时候模仿它，这就是我认为可能不是 10 年后或 40、50 年后，我们的后代将用最复杂的量子机器完成的事情。这太酷了，你知道，因为我猜，而且我的意思是，大多数时候我认为初创公司，初创公司社区以及研究人员都会陷入困境，他们以宏大的愿景开始，但然后他们被事物的经济性所困扰，你知道，比如，让我们先构建东西，构建一个产品，你知道，然后会发生什么，也许然后你就会失去愿景，我并不是说产品不重要，我认为产品是关键，因为我认为，无论是初创公司还是研究人员，你都需要创造可持续的东西。但我认为，你知道，这些疯狂的愿景，进行研究，也许有一天能够瞥见我们是谁，我们为什么存在，我们是什么，你知道，以及这些，这些关于黑洞的宏大问题，我的意思是，你知道，如果我们能够模拟，我认为世界将完全改变，我们将能够做的事情，就像你提到的，我的意思是，时间旅行，无论什么，我认为，因为我认为整个宇宙是量子力学的，如果我们能够模拟大自然，我们将能够构建的那些事情，那些应用将是巨大的，就像举个例子，我认为大约 15 年前，世界上有两个初创公司，我的意思是，也许只是，一个叫 Open，第二个叫 DeepMind，我认为他们开始时有一个宏大的目标，不是创造人工智能来构建应用程序，而是创造研究，并以构建通用机器，一台可能模仿、模拟智能本身的机器为更大的目标。当时他们被嘲笑了，你知道，但现在如果你去看，他们是世界上最有价值的两家公司。所以，所以赞扬你的信念和你的愿景，以及更多的力量，我希望印度有更多的研究人员，你知道，思考宏大，教授，有没有什么量子光子学的突破引起了你的注意，你知道，因为最近我读到关于量子，谈论可能像数百万光子量子比特的目标，然后有来自中国的芯片 X，它构建了一个光子芯片，他们声称可以比这些普通的 Nvidia GPU 快一千倍地推动 AI 任务，你知道，有没有像我这样的光子学突破引起了你的注意？ >> 我看到了两件事，是的，我的意思是，如果你正在看 Psych Quantum 或任何东西，我的意思是，实际上我知道，那个最终成为 Psych Quantum 的团队，作为一个学术研究团队，他们拥有一个知识库，我所说的是创造知识，关于控制光子，学习它们，大约 15 到 20 年的广泛研究工作，当他们决定要创办一家基于这种知识的初创公司时，当然，挑战正如我所说的，相互作用的挑战存在，但他们当时就押注了，即使概率很低，因为我可以产生数千个光子，数百万个光子，我会丢失一些光子，没关系，对吧？因为当我访问它时，我可以随时丢失一些，这就是他们开始的方式，但当公司成立时，当他们开始工作时，他们自己也有知识库，并在此基础上添加，他们正在尝试提出一些我们称之为簇态的东西，或者一些额外的，你知道，新的技术已经被引入，他们正在谈论数百万个量子比特，所以我在大的时候说过，当我们看我们的架构时，无论它是什么，一旦你清楚了我们如何构建门操作，然后光子在移动，并且可扩展性根本不是问题，因为它们对环境稳健，环境不会轻易破坏它们，它们不会像超导或其他硬件那样产生噪音，其中量子比特非常清晰地存在，但噪音却进来了，但这种情况不存在，这就是为什么他们只谈论数百万个量子比特，所以我认为他们当然，正如我所说的，作为一家公司，他们面临着某些挑战，他们有巨额投资，当然，他们也意识到了这些挑战，但我认为他们也在实际地朝着一些可实现的目标努力，但当你谈论突破时，我实际上会说有一些其他的实验，你知道，有一些实验已经完成，以展示经典计算机至今可以做什么，对吧？所以，不久前有一些关于超级计算的说法，但与谷歌不同，但他们实际上已经证明了这一点，但最近他们已经展示了，这太棒了，因为你看到的是一个较少数量的量子比特，但 Genius Group 的那个实验叫做 Boson Sampling 实验。他们用大约 72 个光子进行了实验，平均有 100 多个光子，他们观察了它。但现在，这些实验早在 2009 年就由一些大学在小规模上进行了，他们用 10 个光子进行了实验，然后展示了多个光子如何以某种方式表现，它映射到一些复杂的数学问题。我不会深入研究那个模型问题，但它展示了一些操作，一些，你知道，一些经典计算机无法模拟的结果。我仍然认为这是一个非常好的结果，因为今天它还在继续，它甚至不是通用计算，它只是一个它展示的特定任务。现在我可以想象，如果我能让它通用，我能做什么？我的意思是，有 70 个光子在进入，现在我无法控制它们，它们只是进入，它们与一个特定的基数相互作用，然后你得到一些结果，我只能做一些缺失的事情，填补一些空白，仅此而已，对吧？它本身就可以展示一些世界上没有任何超级计算机可以做的事情，那么如果我能在每个阶段工程化它们，然后观察它，这就是我们所说的通用计算，我可以进行门操作，那么我将拥有，我将打开所有可能性。所以，我认为这是一个非常好的实验演示，但他们花了很长时间，对吧？2010 年是他们第一次展示在非常小的规模上可以做些什么，他们展示了他们花了 8 年时间才建造了那个巨大的设置和结果。所以，这表明有很多希望和承诺，并且在每个中间阶段都有有希望的结果，这表明我知道光子移动得更快，但它们没有记忆，你无法存储它们，但有一些中间的事情，有一些小结果，人们必须尝试它们是否会适合，并且人们正在世界各地努力解决其中一些问题，并且肯定会有解决方案，我坚信这一点，我知道，因为有问题就有解决方案，应该有一些解决方案，但需要多长时间，或者我们应该提出一些其他的论点，一些逻辑上的矛盾来说明为什么你找不到解决方案。我的意思是，如果那是开放的，那么意味着你有机会找到解决方案并观察它。现在问题的可能性是巨大的，当人们谈论药物发现时，我不嘲笑它，这是我们正在谈论的最重要的第一个问题，因为当我们谈论两个或三个粒子时，我们做了一些简单的算法来谈论电子在某些化学物质上，你知道，它可以是苯环或任何这类东西，现在当我们尝试模拟它们时，大约 40 到 50 个原子聚集在一起，那么电子的行为，它们在那里悬挂着，并且有大量的干涉叠加发生，这 40 到 50 个原子，这本身对于任何超级计算机来说都是一个非常困难的问题，可以做到 2 的 n 次方可能性，这意味着，如果我只有大约 50 个原子，然后每个原子的所有电子都可以与任何一个原子相互作用，那么可以产生 2 的 50 次方可能性。这实际上已经达到了最好的经典计算机的极限，我们正在谈论。所以，现在如果我有大约 100 个化学原子相互作用，那么今天没有经典计算机可以模拟它们。人们今天如何进行模拟？量子化学家做了很多模拟。他们是怎么做的？现在他们正在将问题分解成小的子集。他们正在解决它们，然后将它们组合起来。现在当我们分解时，发生的是，当我们谈论量子力学系统中的叠加时，这意味着现在我将问题分解成子集，这意味着我实际上正在切断那些子集之间的量子相互作用，一旦我切断了，我就丢失了大量信息，这意味着我正在近似。所以，人们正在进行的任何化学复合物动力学或模拟，他们都在并行化问题，因为它们非常，没有计算机可以做到现有的计算机。所以，一旦你通过将它们分解成子集并将它们组合起来进行并行化，你就已经进行了大量的近似。现在在这个近似中，一些错误的结果进入了一些有用的东西已经丢失了。现在这就是实际上，这个近似正在帮助我设计针对某个问题的药物，因为那里有一些生物实体。现在我有这个化学物质，它必须反应并解决问题。但是现在我的模拟给了我一些不完全正确的东西。它有它的组成部分，但也有一些不想要的东西。可能缺少一些想要的东西。现在这是一个非常容易理解的问题，人们知道一个非常高效的量子计算机可以有效地模拟大型化学复合物。那么我将获得的作为问题解决方案的任何输出都将是一个精确的解决方案。就像我正在精确地指出我的问题并找到解决方案一样。我不仅仅是在绕圈子。当然，绕圈子有时也能杀死蛇，对吧？但它也会破坏周围的环境。但现在这就是我认为可以解决的最重要的问题，你知道，作为即时影响人类的问题，以及我模拟复杂宇宙的迷人问题，因为这与时间旅行或做任何那些奇妙的技术无关，它只是理解它，它只是让你感到好奇，你知道，坐在这里，我们只做一点点数学，然后尝试理解世界。我们正在谈论宇宙是如何形成的。我们正在谈论黑洞是如何形成的。现在这本身就足够迷人了。现在，如果我能在小规模上模仿它们，那么我们就必须高兴地看到我们已经能够正确地理解大自然的一些东西。对吧？所以，我们可能遗漏了很多东西。并不是说我们理解了，也许它也有助于我们填补量子物理学的空白，你知道，有很多，我们通过实验理解了一些东西，我们正在构建一些我们无法用逻辑观察到的东西，也许是其他东西，来理解它们，看到它们，对吧？所以它打开了一个很大的空间。教授，我的最后一个问题，我的意思是，印度正在推动其贡献，你知道，通过印度国家……你知道，是什么让你对这项倡议最兴奋？以及生态系统在帮助你开发这个……内容系统方面有多支持？这是我的第一个问题。我的第二个问题是，我相信印度有很多开发者生态系统，特别是量子开发者。你对那些想进入量子计算和开发领域的年轻研究者有什么建议？他们从哪里开始？好的。所以，我认为第一个问题，我认为我们已经做得非常出色，国家量子任务。事实上，我应该说，在获得国家量子任务之前，我认为 DST 已经有一个试点模型，称为 QUEST，量子赋能科学与技术，你知道，我们称之为小型试点项目，给了大约 51 位研究人员。我也参与其中。我当时也在研究理论方面。所以，那时我们都是国家量子任务初稿的贡献者。我知道首先是学术界。实际上是学术界被要求提出可能的问题，然后这份草稿后来被政府组织，如 DRDO、空间组织等采纳，以提供他们的意见，包括最终用户，如军队和国防。所以他们也提供了一些他们想要的东西的意见。所以，这是一份非常集体的文件，后来成为国家量子任务文件，而且它确实，我的意思是，没有人说没有参与学术界，所有政府基金会都参与了，以提出印度应该如何展望。这是我们所有人的贡献，我应该说，人们，我的意思是，当然我们有数十亿人，但工作的人，大多数人都发挥了作用。现在执行，它确实发生了，你知道，好吧，因为在某个时候有宣布，然后你知道，我们必须遵循某些规则，它不能只是说，好吧，这里有一些钱，你拿去玩，然后你玩弄它。这是公共资金，并且有一个必须遵循的课程，如何分配它们，一切都是可追溯的，对吧？你如何遵循它，但最好的事情是去年宣布了呼吁，对吧？所以，尽管有人在努力，但每个人都必须竞争。所以，去年年初，二月、三月，有了呼吁，我们被告知了某些时间表，然后我们被告知八月、九月将做出决定，谁将领导某些项目，令人惊讶的是，而且是好的，它按时发生了，这是因为这是一项艰巨的任务，要找到审稿人，评估数百名提交者，因为有很多提交，每个人都必须被评估，预算必须被考虑，最终它在去年获得批准，并在今年年初的二月、三月，我们开始获得一些资金，而且非常，现在我认为我们已经获得了第一年的资金，我们所要求的，我们第一年所需要的，因为采购，有某些东西，生态系统非常重要，就像你知道的，我也将你的第二个问题带到这里，就像你知道的，如果你说有一些资金，你必须建造，那么你就不能试图从零开始建造一切，然后说我要做经济。我们首先需要，你知道，重新评估，你知道，首先建造现有的东西，我们可以采购的东西，然后建造某些东西，然后开始在基层建立生态系统，让初创公司，行业，其他政府组织参与进来。所以，生态系统必须建立，而所有这些都已在国家量子任务下规划，即量子计算中心，或 IIT Madras 中心，或孟买中心，或德里中心，有四个中心已经建立，它们已被委托或被要求现在关注初创公司生态系统，其他小型参与者的参与，以便将它们全部聚集起来，为我们正在构建的更大的系统构建一些必需的较小组件。所以，这一切都已开始得很好，而且有资金，让我们开始，正如我们所要求的。当然，采购，我们必须再次遵循政府程序，对吧？所以，投标必须完成，并且当有公共资金时，有问责制，我们必须遵循规则，有一些小的延迟，但那没关系，应该是这样的，对吧？你知道，所以一切都完好无损，所以，是的，我们已经，自从我们获得资金以来，我们已经建造了一个新的实验室，实际上，因为我们之前的实验室很小，我们有我们的量子比特，现在由于我们必须从小规模扩展到更大数量的量子比特，我们需要更大的空间，以便我们可以建造，所有这些都已到位，以便继续进行。所以，我应该说，至少在资金方面，它非常支持，我们实际上，当我说是学术界参与时，DST 的人与我们密切合作，所以我们现在已经有一年了，自从我们在多个层面进行互动以来，它已经成为你的一部分，一个群体，你知道，我们有自己的分歧，我们有自己的争论，我们有时对某些事情不满意，他们对某些事情不满意，你知道，因为正如你所说的，当有公共资金时，他们的期望，你知道，不能是他们需要可交付成果，对吧？他们可以通过这种方式进行追溯，当资金来自公共资金时，你向我们展示可交付成果，我们向你展示已使用的产品，而这可能就是我们所说的，当然，基础科学非常重要，我们也需要探索。现在我们处于中间部分，你知道，他们不阻止我们做我们的科学。他们不阻止我们做，但当然，他们期望我们做一些可交付成果。现在，这是我们作为研究人员的任务，看看我们如何平衡，因为我们构建知识库的长期利益不应与交付中间可交付成果相抵触，这些可交付成果对于说服政府和公众来说是必要的，你知道，他们正在看到量子将改变，有一些产出，可以被利用。现在谈到生态系统，我认为有很多，你知道，“量子”这个词被炒作了，对吧？每个人都在谈论它，每个人都会说我将，但这是，为什么人们想要开始应该非常谨慎。这不仅仅是把现有的东西放在一起，或者只是在现有的东西上添加一些额外的软件。它仍然很深，你知道，当学生有时来找我说，我尝试看一些软件，尝试看一些算法，我有一些专业知识，我需要知道首先你也要了解背景中发生的事情，因为正如我所说的，没有确定的架构，没有确定的硬件，这意味着你需要了解背景是什么，只有这样，任何小的变化你才能适应。如果一个初创公司今天来，没有，你知道，没有如何适应时间的愿景，你知道，我可以这样说，我建造，例如，如果我说我建造了一个特定的设备，它会产生一些随机数，然后我说，如果我不考虑如何小型化，如何使用新的量子资源，或者你知道，新的量子思想进入系统，那么我就死定了，因为别人会来扮演重要角色。所以他们必须进入这个领域，带着长期的承诺和愿景，那么肯定会有，因为知识库，即使是初创公司拥有的，也是不可替代的。没有人会，你知道，可能在 10 年后，5 年后，有人可能会带着巨额投资来，但他们仍然没有人们随着时间的推移积累的所有知识。所以，这应该是重点，你知道，他们可以有一个初创公司，他们可以，他们可以建造一些小的组件，无论是什么，但他们应该开始像在那里建立一个小研发实验室一样思考，这意味着他们有一些相关的研究，而这就是初创公司能够长期成为重要参与者的知识库。>> 对，对，教授，我非常感谢您抽出时间参加播客，正如我之前提到的，我认为我们正处于一个绝佳的时刻。我认为未来 10 到 15 年将是超级棒的。你知道，我们，整个，我的意思是，计算的范式是零和一。但我们在这里。我的意思是，你知道，我们正在过渡，我们将进入那个特殊的地方，你知道，就像另一个地方一样，你知道，我们深入到，你知道，用量子比特计算，你知道，生命的源代码，你知道，就像我们都是由什么组成的。我认为，多年来，你知道，印度一直在软件方面领先，我希望，你知道，它能够站起来，并在我们正在进入的这个新领域，你知道，量子计算领域中被计算在内。我认为，你知道，政府的推动，以及像您这样的研究人员，在构建这些新事物中发挥作用，赞扬你们所有人，我希望有更多的合作，因为我，以前，你知道，所有的突破都只发生在美国或中国，你知道，但现在，我认为世界正在改变，你知道，印度也已经被计算在内，你知道，所以我希望这种势头能够继续下去，我们非常期待看到下一个量子光子技术突破来自印度，你知道，由印度赋能，你知道，可能来自你的实验室，你知道，所以真的，真的感谢您抽出时间参加播客，对我来说，听众，如果你喜欢你在这里看到和听到的，请按下订阅按钮，下次再见。谢谢。谢谢，兄弟。真的非常感谢。